Economics of Transition
Volume 8 (3) 2000, 577–601
The limits of discipline
Ownership and hard budget constraints in the
1
transition economies
†
‡
Roman Frydman*, Cheryl Gray**, Marek Hessel and Andrzej Rapaczynski
*Department of Economics, New York University, 269 Mercer Street, New York, NY 10003, USA and
Privatization Project. Tel: + (1) 212 9987 8900; E-mail: [email protected]
**The World Bank
†
Formerly of Graduate School of Business, Fordham University, and Privatization Project
‡
Columbia University School of Law, and Privatization Project
Abstract
The existing literature on soft budget constraints suggests that firms may be subsidized for
political reasons or because of the creditors’ desire to recover a part of the sunk cost
invested in an earlier period. In all these models hard budget constraints are viewed as
being, in principle, capable of inducing the necessary restructuring behaviour on the level
of the firm. This paper argues that the imposition of financial discipline is not sufficient to
remedy ownership and governance-related deficiencies of corporate performance. Using
evidence from the post-communist transition economies, the paper shows that a policy of
hard budget constraints cannot induce successful revenue restructuring, which requires
entrepreneurial incentives inherent in certain ownership types (most notably, outside
investors). The paper also shows that the policy of hard budget constraints falters when
state firms, because of inferior revenue performance and less willingness to meet payment
obligations, continue to pose a higher credit risk than privatized firms. The brunt of state
firms’ lower creditworthiness falls on state creditors. But the ‘softness’ of these creditors,
while harmful in many ways, is not necessarily irrational, if it prevents the demise of firms
that are in principle capable of successful restructuring through ownership changes.
JEL classification: G32, P17, P27, P31.
Keywords: ownership, financial discipline, performance, transition.
1 The authors would like to thank Joel Turkewitz for his contributions to the design and implementation of
the survey instrument, and Mihaela Popescu for her extraordinary assistance in the analysis of the data.
The authors also thank Sarbajit Sinha for computer support in the initial stages of research. Helpful
conversations with, and comments from, an anonymous referee, Marvin Chirelstein, Simeon Djankov,
William Greene, Glenn Hubbard, Edmund Phelps, Mark Roe and participants in the faculty workshop at
Yale Law School are also gratefully acknowledged. The authors are grateful to the CEU Foundation, the
Open Society Institute and the World Bank for supporting research on this paper. CV Starr Center for
Applied Economics at New York University has provided additional support for Roman Frydman’s
research. None of these institutions are responsible for the opinions expressed in this paper.
©
The European Bank for Reconstruction and Development, 2000.
Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA.
578
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
1. Introduction
Financial discipline is often viewed as the most important prerequisite of efficient
corporate performance, and indeed of the health of the economy as a whole. ‘Soft
budget constraints’ have long been seen as a primary source of inefficiency of
communist firms and as being responsible for the ultimate failure of the socialist
regimes (Gomulka, 1985; Kornai, 1993). More recently, financial crises around the
world have been attributed to lax financial practices on governmental, bank and
firm levels (Mitchell, 1999, 2000). Not surprisingly, therefore, the various ‘rescue
packages’ have stressed the need for greater financial discipline and made it one
of the most important conditions of economic assistance.
The existing literature on soft budget constraints suggests that firms may be
subsidized for political reasons, such as the desire of officials to subsidize
employment (Kornai, 1993, 1998) or enlarge their political constituency (Shleifer
and Vishny, 1994), or because of the creditors’ desire to recover a part of the sunk
cost invested in an earlier period (Dewatripont and Maskin, 1995; Maskin and Xu,
1999). In all these models, it is assumed that firm behaviour is a function of the
constraints faced by the firm and that the hardening of financial discipline
introduces the preconditions of efficient operation. Hard budget constraints are
thus seen as being, in principle, capable of inducing all necessary restructuring
behaviour on the level of the firm. Privatization, in this context, is understood as a
precommitment device helping the state make credible its promise not to continue
subsidies in the future.
We do not question that financial discipline plays a very important role in
corporate performance and that hard budget constraints restrict waste and force
better cost management. But we argue in this paper, using empirical evidence
from the post-communist transition economies, that there are clear limits,
deriving from the governance and ownership structure of firms, to what financial
discipline can accomplish. In particular, the concept of restructuring implicit in
the standard models of the impact of financial discipline focuses on cost efficiency
and is not sufficiently robust to include other preconditions of a firm’s ultimate
success (and the vibrancy of the economy as a whole), namely the degree of
inventiveness, creativity, and readiness to accept risk on the part of the persons
responsible for corporate decisions. These characteristics, often embraced in the
general concept of entrepreneurship, have only an indirect relation to financial
discipline, and there are reasons to believe that financial discipline is not a
sufficient condition for their promotion and development. Moreover,
privatization, when properly designed, is not just a precommitment device
assuring truly hard budget constraints in the future, but rather a form of firmlevel restructuring necessary for the unleashing of the spontaneous
entrepreneurial energies inherent in certain forms of ownership. Without such
ownership reforms, even in the presence of hard budget constraints, postcommunist firms are incapable of generating sufficient revenues to be able to
service their financial obligations.
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
579
Our evidence comes from a study of small and medium-sized firms in the
early stages of the transition in the Czech Republic, Hungary and Poland. In all
three countries, trade and bank creditors were imposing harder constraints on
loss-making firms (Schaffer, 1998). As a result, not only private but also state
firms were forced to economise on available financial resources (Pinto, Belka and
Krajewski, 1993). Extensive employment changes among the largest state firms in
Central Europe, documented by Balcerowicz, Gray and Hashi (1998), were also
presumably in large part prompted by the tightening of budget constraints. The
streamlining of the cost side of the state enterprises’ operations in response to
external discipline was in fact quite effective: the restructuring of state enterprises
in the transition environment of Central Europe tended to lower their costs by
margins similar to those achieved in privatized firms (Frydman et al., 1999;
Frydman, Hessel and Rapaczynski, 1999).2
The impact of hard budget constraints is very different on revenue
restructuring. What ultimately matters in credit markets is a firm’s ability to pay,
and even the strictest financial discipline will not allow firms to keep covering
their obligations if they cannot generate sufficient revenues. We have argued
elsewhere (Frydman et al., 1999) that the ability to generate revenue, especially
during periods of rapid economic change, is inherently tied to firm ownership,
and that firms privatized to certain types of owners (primarily outside investors)
enjoy a significant performance advantage in this respect over firms controlled by
corporate insiders (management or employees) or the state. In this paper, we
provide further evidence of the connection between ownership and revenue
generation by showing that the marked difference in revenue performance
between privatized firms owned by outside investors and those controlled by
insiders are not due to differences in financial discipline, since the constraints
faced by both insider- and outsider-controlled firms are equally ‘hard’.
The inability of firms owned by corporate insiders or the state to match the
revenue performance of firms owned by outside investors means that they are
less able to repay their debts and constitute a higher credit risk to their lenders.
But in addition to the higher likelihood of default due to the lower ability to pay,
state enterprises (but not insider-controlled firms) are also less likely to honour
their obligations than other types of firms with a similar ability to pay, and we
trace this aspect of state firms’ behaviour to the attitude of their state creditors.
Our findings also shed light on the behaviour of different types of creditors in
the early stages of the post-communist transition. In well-functioning credit
markets, differences in creditworthiness of firms should be reflected in the
2 The effects of the hardening of budget constraints on enterprise and bank performance in various
transition economies have been analysed by, among others, Coricelli and Thorne (1993), Coricelli and
Djankov (1998), Bonin and Schaffer (1995), Berglof and Roland (1997, 1998), Bai and Wang (1998), Djankov
(1999), and Perotti (1998). The literature on soft budget constraints in the context of post-communist
restructuring has been reviewed by Djankov and Murrell (2000). In other contexts, budget constraints have
been analysed as an essential factor in the functioning of financial systems (Dewatripont and Maskin, 1995)
and federalism (Qian and Roland, 1998).
580
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
allocation of credit. One would therefore expect the higher credit riskiness of state
firms to result in a stricter rationing of credit to those firms. Our data suggest this
is indeed the case for trade credit. However, state banks and tax authorities
appear significantly ‘softer’ in imposing financial discipline on state firms than on
their privatized and private counterparts.
To be sure, the fact that state creditors continue to impose laxer discipline on
state firms effectively amounts to a new subsidy. But this subsidy need not result
from a merely undesirable kind of politicization of credit decisions that a better
implementation of a tough credit policy would perhaps be able to avoid. Unless
one is willing to accept a likely demise of a large number of state firms that,
despite their inability to generate sufficient cash flows, are nevertheless known to
be potentially valuable, ‘rock hard’ budget constraints might simply be
unrealistic, and the subsidy provided by state creditors might, to some extent, be
rational, since the firms in question are probably capable of increasing their
revenue generation through ownership transformations (Frydman et al., 1999).
The conclusion we draw from this is not that the policy of hard budget constraints
should be relaxed or abandoned, but that it cannot be made truly effective unless
accompanied by speedy and effective privatization.
2. The sample
The analysis presented here is based on a survey of mid-sized manufacturing
firms in the Czech Republic, Hungary and Poland conducted in 1994–95. The
firms were classified into new private, privatized, and state firms.3 The privatized
firms were further divided into those owned by insiders (management or
employees) or outside investors (foreigners, domestic individuals, domestic
financial and non-financial companies), and state firms were divided into noncorporatized and corporatized entities. The subsample of firms used in this paper
consists of 216 firms, 70 of which were in the Czech Republic, 90 in Hungary, and
56 in Poland. Thirty-one per cent of sample firms were held by the state (17 per
cent as corporatized entities), 43 per cent were privatized (19 per cent to insiders
and 81 per cent to outsiders), and 26 per cent were private firms. At the time of
the survey, state and privatized firms were quite similar in terms of their size,
with the average employment of about 580 full-time employees and the annual
sales just above US$ 12.5 million; private firms were smaller, with average
employment of about 200 employees and sales of about US$ 7 million.
(Additional sample description can be found in the appendix.)
3 A firm was classified as private if it was never a state-owned enterprise. A previously state-owned firm
was considered privatized if the combined holdings of private parties gave them a blocking power over
major company decisions. In the absence of such power, the firm was classified, depending on its legal
form, as a non-corporatized or corporatized state enterprise. Given very high ownership concentration (See
the appendix), the ownership of a privatized firm was identified with that of the largest private owner.
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
581
Table 1. Ownership, revenue performance, and short-term credit arrears
Corporatized
State
firms
enterprises
Median annualized rate of
revenue growth* (%)
Percentage of firms with
declining revenues**
Overall (1990–93)
1990–91
1991–92
1992–93
Median level of arrears (%
of 1993 revenues)***
Percentage of firms in
arrears (1993)
Overall
to bank creditors
to trade creditors
to tax authorities
–24.0
–16.7
Insiderowned
privatized
firms**
–7.8
96.7%
87.5%
82.8%
90.0%
31.5%
89.5%
94.7%
71.4%
86.8%
25.6%
66.7 %
100.0%
100.0%
61.1 %
19.8%
52.0%
63.6%
57.7%
54.7%
4.8%
38.2%
43.5%
51.4%
45.5%
3.1%
80.0%
70.4%
71.4%
37.7%
65.5%
41.7%
36.7%
31.6%
20.0%
11.8%
12.5%
16.7%
35.5%
20.0%
23.9%
10.7%
32.6%
27.5%
4.3%
18.2%
OutsiderPrivate
owned
(nonprivatized privatized)
firms**
firms
–2.0
18.8
Notes: * For state firms, the rate of revenue growth annualized over 1990–93 period; for privatized and
private firms, the rate annualized over the period a firm operated as a private entity; ** For privatized
firms, post-privatization periods only; *** Firms in arrears only.
Table 1 illustrates the revenue performance of sample firms over the 1990–93
period and summarizes overdue short-term credit obligations (short-term bank
credit, taxes to central and local governments, and trade credit) among firms with
different ownership types. Although a majority of firms in our sample suffered
revenue losses in every year between 1990 and 1993, the main fault line in
revenue decline is between state and private firms. As of the end of 1993, 96.7 per
cent of state-owned (non-corporatized) enterprises, 89.5 per cent of corporatized
enterprises, 54.8 per cent of privatized firms, and 38.2 per cent of private (not
privatized) firms had lower revenues than in 1990.
The decline of revenues in the early years of transition was bound to put many
firms in financial difficulties. Overall, 45 per cent of the firms in our sample were
in arrears to one or more of their short-term creditors.4 Not surprisingly, the
4 In evaluating credit performance, we considered a firm to be in bank credit or tax arrears if any of its
payments were overdue. For trade credit, we defined overdue payments as those late by 60 or more days.
This extension of the usual repayment period was necessary because delays in repaying trade credit within
582
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
percentage was markedly higher among firms with declining revenues: 57 per
cent of such firms were in arrears to one or more of their creditors, as compared to
24 per cent of those with growing sales.
State firms dwarfed all other types of firms in failing to meet their financial
obligations to all types of creditors: Nearly 4 out of every 5 non-corporatized state
enterprises were in arrears to one or more of their short-term creditors, as
compared to 2 out of 3 of corporatized state firms, 1 out of 3 privatized and 1 out
of 7 private firms. Moreover, among the firms in arrears, the median level of
combined arrears in state firms, amounting to nearly 32 per cent of annual
revenues, was higher than among the corporatized (25.6 per cent), privatized (4.9
per cent) and private (3.1 per cent) firms. (Insider-owned privatized firms had
somewhat higher arrears than other private firms, but the statistical significance
of this is undermined by the fact that very few of them were in arrears to begin
with.)
3. Ownership and performance
The effects of ownership on firm performance, which may be glimpsed from the
statistics in Table 1, were analysed in detail in Frydman et al. (1999). Using fixedeffects panel regressions to estimate the impact of ownership type on firm
performance, Frydman et al. (1999) found that outsider-owned privatized firms
had significantly (nearly 10 percentage points) higher annual revenue growth
than either state or insider-owned firms, but that no significant differences could
be detected in terms of (labour or material) cost performance among any two
groups of firms (see Table 2). The performance of state and insider-owned firms
was statistically indistinguishable, both in terms of revenue and cost performance,
as was the (not reported in Table 2) performance of corporatized and noncorporatized state enterprises.5
30 days were quite common during the early stages of transition, particularly in the Czech Republic and
Poland. In both of these countries, one out of every two firms failed to repay its suppliers within 30 days,
the ratio dropping to one in three past 60 days (in Hungary, the ratio stood at one in four irrespective of
the time period.) Moreover, the widespread incidence of ‘short’ delays (under 60 days) overshadowed any
systematic ownership differences, which surfaced only with the incidence of more persistent delays.
5 The performance comparisons in Frydman
et al. (1999) included only state and privatized firms; since new
private firms had not been burdened with the communist heritage and did not face the same challenges as
the former state firms, they were excluded from the comparisons. The performance of corporatized state
firms did not differ in any relevant way from that of other state firms, and the two types of state-owned
firms were therefore not grouped separately. In the estimates of creditworthiness of different types of firms
in this paper, however, new private firms are included and corporatized state firms are treated as a
separate grouping. For recent analysis focusing on the financing of new private firms in the transition
environment, see McMillan and Woodruff (1999) and Bratkowski, Grosfeld and Rostowski (2000).
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
583
Table 2. The effect of privatization for outsider- and insider-owned firms
Annual rate of growth of Annual rate of growth of
revenue, 1990–93
cost per unit of revenue,
1990–93
Privatization effects
1
Outsiders
1
Insiders
Test statistics for the model
Test statistics for the equality of
group effects
9.70*
(3.64)
0.68
(5.28)
no. of obs=513
F=7.05*
adj R2=0.13
F= 1.24
p=0.29
–4.36
(3.33)
1.12
(4.45)
no. of obs=347
F=5.27*
adj R2=0.14
F= 0.42
p=0.66
Notes: (1)A dummy variable set to 1 for the post-privatization performance of privatized firms where the
given type of owner is the largest shareholder, 0 otherwise. State-owned enterprises are the reference
group. * p 0.01, ** p 0.05, *** p 0.10. Standard errors in parenthesis, significant coefficients bold-faced.
Group-effects estimates. Number of firms = 218. Group effects and coefficients on initial values of
performance measure and country-year variables not reported here. (Full statistics are reported in
Frydman et al., 1999.)
These results were controlled for differences in the macroeconomic
environment of the three countries; they persisted across all industrial sectors in
which the sample firms operated, and could not be attributed to possible
performance ‘dips’ in the years immediately preceding privatization. They were
not driven by either a handful of well-performing firms or a few poorly
performing firms. We have also shown in Frydman et al., (1999) that no selection
bias was involved in the process of choosing state firms for privatization, and in
particular that better performing firms were not selected for privatization to
outside investors.
4. Creditworthiness and the ability to pay
Is it possible that the differences in performance observed in Table 2 could be
explained by the fact that firms with different types of owners faced different
budget constraints? Did creditors give special treatment to certain types of firms
and was that treatment responsible for the revenue behaviour of the firms? We
begin this inquiry by looking for the determinants of the likelihood that a firm
with certain characteristics (such as performance, ownership type, etc.) would
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
584
have defaulted on any of its outstanding short-term credit obligations to any of its
creditors during the 1990–93 period. The higher the probability that a firm would
have been late in meeting its obligations to one of its creditors, the lower the
firm’s overall creditworthiness and the higher the credit risk it represents to that
creditor. The degree to which a firm’s creditworthiness determines its access to
finance is, in turn, an indication of the hardness of its budget constraints.
Table 3. Probability of default on obligations to different types of creditors, 1990–93
Creditor to whom the debt is due
Variable
(1)
Bank
creditors
(2)
Tax
authorities
(3)
State (banks
or tax
authorities)
(4)
Trade
creditors
Constant
–0.60*
(0.08)
–0.011*
(0.003)
0.000
(0.000)
0.43*
(0.13)
0.04
(0.10)
–0.26
(0.17)
–0.01
(0.11)
0.30*
(0.10)
0.35*
(0.11)
n=209
2
χ = 70.08*
–0.67*
(0.07)
–0.008*
(0.002)
0.000
(0.000)
0.38*
(0.10)
0.06
(0.08)
0.001
(0.11)
0.05
(0.08)
0.35*
(0.08)
0.38*
(0.09)
n=216
2
χ = 77.83*
–0.58*
(0.09)
–0.013*
(0.003)
0.000
(0.000)
0.33*
(0.14)
0.04
(0.11)
–0.25
(0.16)
0.01
(0.11)
0.35*
(0.10)
0.30*
(0.11)
n=201
2
χ = 55.69*
–0.21*
(0.06)
–0.006*
(0.002)
–0.000
(0.000)
0.08
(0.10)
0.08
(0.09)
–0.09
(0.13)
–0.30*
(0.11)
–0.12
(0.08)
–0.01
(0.09)
n=203
2
χ = 38.96*
Rate of growth of revenues for
firms with decreasing revenues1
Rate of growth of revenues for
firms with increasing revenues1
State firms2
Corporatized state firm3
Insider-owned privatized firm4
Private (non-privatized) firm5
Hungary6
Poland7
No. of firms (n)
Notes: *p 0.05 **p 0.10; significant coefficients bold-faced. Maximum likelihood estimates of binomial
probit models.
(1) Revenues are measured in constant local prices; rates of revenue growth are annualized over the 1990–
93 period for state firms and over the post-privatization period for privatized firms; (2) A dummy variable
equal to 1 if the firm is state, zero otherwise; (3) A dummy variable equal to 1 if the firm is corporatized,
zero otherwise; (4) A dummy variable equal to 1 if the firm is privatized, insider-owned, zero otherwise;
(5) A dummy variable equal to 1 if the firm is a newly-founded by private parties, zero otherwise; (6) A
dummy variable equal to 1 if the firm is in Hungary, zero otherwise; (7) A dummy variable equal to 1 if the
firm is in Poland, zero otherwise.
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
585
Table 3 presents the results of a probit model estimating the likelihood of a
firm’s default on its short-term obligations as a function of the firm’s revenue
performance, ownership type, and the country in which the firm is located.6 We
use the annualized rate of revenue growth to measure the firm’s revenue
performance.7 Since falling revenues are apt to affect the likelihood of default
differently from rising ones, we split the revenue growth into two corresponding
variables. We estimate separately likelihoods of defaults on obligations to the
following types of short-term creditors: (1) bank creditors; (2) tax authorities;
(3) any state creditors (banks or tax authorities); and (4) trade creditors.
(Throughout, privatized firms owned by outside investors in the Czech Republic
serve as the reference group).
Note first the consistent relationship between creditworthiness and the rate of
declining revenue growth in Table 3. In our interpretation, the reason why the
coefficients of this variable are highly significant throughout all four regressions is
that the declining rate of revenue growth stands (inversely) for a firm’s ability to
meet its financial obligations. Clearly, a number of other factors, such as cash flows,
the excess of sales over costs, and the degree of leverage, might also affect a firm’s
ability to pay, so that singling out declining revenue growth as the proxy may be
controversial. To some extent this choice is dictated by the fact that information
concerning the other factors, such as cash flows, cost of capital, and profitability,
was notoriously unreliable in the early stages of the post-communist transition.
We have only limited data on those factors, and we doubt that creditors were able
to rely on them to any significant extent.
But we have a more important reason for treating a firm’s ability to service its
debt obligations in the environment of the early post-communist transition as
primarily a function of its ability to generate revenues. We have seen in Table 2
that different types of firms in the early 1990s differed most not in their ability to
control costs but in their capacity to restructure their products and find new
markets, so as to achieve a level of sales capable of assuring their viability. Once a
firm’s revenues were growing, costs could be kept in check, ensuring in most
cases that the firm would be able to meet its payment obligations % whether or not
it actually met them depended then mostly on other factors, and the rate of
revenue growth was not by itself predictive of the likelihood of running arrears
(which explains its insignificance when the revenues were rising). When revenues
were collapsing, however, even the most stringent cost cutting measures were
unlikely to provide sufficient savings to enable the firm to service its debts.
Indeed, while we see the close relation between the inability to generate new
6 We also tested for the impact of any sectoral differences among firms by including two-digit industry
dummies as well as combinations of such dummies (capturing the difference between consumer and
industrial goods sectors) among the variables in our model. We found that all such sectoral effects lacked
significance and their inclusion left the estimates of other variables virtually intact. Including size of firms
in the regressions also did not affect the results.
7 If R denotes a firm’s performance level in year t, the annualized rate of revenue growth, r, is an imputed
t
rate which satisfies PT/Pt = (1 + r )(T – t ) over the appropriate time interval (T – t), T > t.
586
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
revenues and lower creditworthiness, we could discern no statistically significant
relation between a firm’s cost performance and the likelihood of its defaulting on
payment obligations.8
It is important to note that if the rate of falling revenues is indeed a proxy for a
firm’s ability to meet its financial obligations, then this very fact, quite
independently of the coefficients on the ownership dummy, means that state and
insider-controlled firms were systematically less creditworthy than their outsideinvestor-controlled counterparts.9 This is because in the early years of the
transition, state and insider-controlled firms lost revenues much faster than firms
that had either never been state-owned or had been privatized to outsiders:
indeed, over 9 out of 10 state and corporatized firms, and 2 out of 3 insidercontrolled firms in our sample had declining revenues over the sample period.
5. Willingness to pay and soft budget constraints
A firm’s decision to repay any individual obligation, however, depends on more
than its ability to come up with the cash; it involves a decision on whether to
allocate any available cash to this or another purpose. This decision reflects a
component of the firm’s creditworthiness that goes beyond the ability to repay:
we assume that decisions to repay creditors are economic in nature, and that they
reflect the perceived balance between the costs of delaying payments to some or
all creditors and the costs of other foregone opportunities that a timely payment
entails. We will refer to this decision as reflecting the firm’s willingness (or
propensity) to pay, and we gauge it by the differential likelihood of its defaulting
on its debt payments, as compared to other (types of) firms with a similar ability
to pay.10 In other words, we interpret the significance of the coefficient of the
8 Among the tests we performed, we substituted analogous cost variables for the revenue variables in the
equation reported in Table 3. We found all coefficients on these variables to be insignificant, except for one
anomalous result in the case of trade creditors, where lower costs were weakly predictive of the higher
probability of default. When both cost and revenue variables were included in the same equation, data
limitations (smaller number of firms and greater degrees of freedom) did not allow us to obtain statistically
significant results in the case of certain types of creditors. But numerically the results were always
consistent with the hypothesis that creditworthiness is dependent on revenue declines and independent of
cost performance.
9 Although new private firms were not included in the regressions reported in Table 2, their average rate of
revenue growth was still higher than that of all types of privatized firms.
10 This way of gauging the differences in the willingness or propensity to repay depends, of course, on the
closeness of ‘fit’ between revenue growth and the ability to repay. To the extent that revenue growth may
not be a perfect proxy, and thus may not capture certain aspects of a firm’s ability to repay, those aspects
may be ‘picked up’ by other variables in our equations, especially those pertaining to ownership type. The
most important potential candidate for such leakage would be the state firms’ inferior cost performance.
But, as we have explained already, there is no statistically significant relation between cost performance
and creditworthiness in our sample and state firms do not, as a matter fact, perform worse on costs than
privatized companies.
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
587
ownership dummy in the equation reported in Table 3 as indicating that the firms
of the ownership type involved are less willing or prone to repay their debts than
the reference group of privatized firms controlled by outside investors.
A firm’s ability to repay may be a function of the macroeconomic environment
(which has an impact on overall demand) and firm-specific characteristics (in our
model, the ability to generate sufficient revenues to meet payment obligations).
The same firm’s willingness or propensity to repay is, in turn, a function of three
different types of factors: (1) system-specific characteristics of the legal and
economic environment, such as the extent of credit market development, the
nature and effectiveness of bankruptcy laws, etc.; (2) firm-specific characteristics,
corresponding to how the firm values its relation with a given creditor relative to
other claimants (including its labour force and any stakeholder constituencies that
might support the priority of some claimants over others) and other available
investment opportunities; and (3) creditor-specific characteristics, related to how
stringently the creditor will react to non-payment or even a possibility of default.
The system-specific determinants are exogenous and affect many firms equally.
The second factor may reflect firm-level politicization of business decisions, such
as the difficulty that some state firms, especially those in which worker councils
have significant powers, may have with laying off employees or keeping salaries
in check, or the pressures from state officials interested in furthering various
political objectives at the expense of the firm’s standing with its creditors. The
third factor is of special importance here because it reflects a potential softness of
the firm’s budget constraints. Moreover, the third factor clearly affects the second,
as the softness of the budget constraint is, in turn, anticipated by the firm and
becomes reflected in the firm’s priorities with respect to the allocation of its
financial resources.
Although the lower willingness of some firms to repay their debts can thus
have a number of explanations, it is the sole channel, in our model, through which
the relative hardness of the budget constraint faced by different types of firms can
be gauged. In other words, if some or all creditors are less strict in demanding
repayment from firms with a certain ownership type, the firms of this type will be
more likely than others to default on their obligations, over and above what could
be expected from looking only at their ability to pay. This increased likelihood %
which we identified with the lower willingness to repay % will, in turn, show in
the significance of the coefficient of the ownership dummy for the type of
ownership in question. Conversely, if the coefficient of the ownership dummy is
not significant, it seems safe to assume that the firms of that ownership type do
not receive preferential treatment from their creditors, as compared to the
reference group of firms (outsider-owned privatized firms in Table 3).
A number of results in Table 3 deserve special notice in the light of this
interpretation. The first is the coefficient of the ownership dummy for insiderowned privatized firms, which is not significant in all four equations. This
indicates that insider-controlled firms do not lag behind the reference group in
their willingness to repay their debts; if anything, the size of coefficient of the
588
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
ownership dummy for insider-owned privatized firms suggests that insiderowned firms may be more likely than their outsider-owned counterparts to meet
their obligations.11 These firms thus do not seem to face any softer budget
constraints than outsider-controlled firms. (In fact, with one exception % the
treatment of new private firms by trade creditors % to be discussed later, all types
of private firms, whether new or privatized, as well as corporatized state firms,
seem to be in the same boat as far as their treatment by all types of creditors is
concerned.) Recall, however, that in terms of their performance, insider-owned
firms are not significantly different from state enterprises, and that they lag very
significantly behind outsider-owned privatized firms in terms of their revenue
generating ability. The combined presence of hard budget constraints and the
inferior revenue performance of insider-owned firms strongly suggests that
certain aspects of firm performance – those involving risk taking and
entrepreneurship % are not induced by financial discipline, and that the presence
of hard budget constraints cannot substitute for effective ownership reforms.
The same conclusion follows from the fact that the coefficients of the
ownership dummy are not significant for state-owned corporatized companies in
all four equations in Table 3, since these companies also did not differ in any
significant way from other state companies in terms of their performance (and
revenue performance in particular). Again, the fact that the budget constraints
faced by corporatized state firms were as hard as those faced by all types of
privatized firms does not seem to have been sufficient to bring the crucial revenue
performance of corporatized firms any closer to that of privatized firms owned by
outside investors.
Another result in Table 3, to be viewed in conjunction with the presence of
hard budget constraints facing all types of private firms, is the significantly
positive coefficient of the ownership dummy for non-corporatized state
enterprises across three of the four equations. According to our model, this
significance indicates that, in addition to being less able than outsider-controlled
privatized or new private firms to meet their financial obligations, noncorporatized state enterprises were also less willing to meet them. In other words,
even after their ability to pay (as measured by the rate of revenue growth) was
controlled for, these state firms were still more likely to default on their
obligations to all of their lenders except trade creditors.
While softer budget constraints could be one of the explanations of the lesser
willingness of non-corporatized state firms to meet their financial obligations to
most creditors, we have seen that other factors, such as a pressure to avoid layoffs or other political objectives, may also be responsible. To what extent the
greater probability of default by state firms may be due to the softness of their
creditors will be the subject of further analysis in Section 7 below.
11 The contrast between the insider- and outsider-owned firms becomes significant if all the creditors are
combined in one equation.
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
589
6. Obligations to state and private creditors
In a well-functioning market, and in the absence of any distinctive seniority
structure of outstanding short-term obligations, the likelihood of a firm’s being
late on payments to particular types of creditors should match the ‘overall’
creditworthiness of the firm. Although we have no information on the terms of
credit granted by different types of creditors, we see no prior reason for any shortterm debts to be subordinated to other short-term debts in a systematic fashion.
We therefore assume that, in principle, all short-term credit obligations are to be
met pari passu and we interpret systematic differences in the likelihoods of arrears
to different types of creditors as evidence that some firms give and/or receive
differential treatment to or from some types of creditors. Which creditors the
embattled firms tended to pay first and which they tended to skip when money
was short reveals quite a lot about the effectiveness of financial market reforms.
Noteworthy in this context is the fact that non-corporatized state enterprises
that appear less willing to repay their short-term debts to banks and tax
authorities are nevertheless no less ‘resolved’ than their privatized or
corporatized counterparts to pay their trade creditors. This may be due to several
factors. Trade creditors may have found it easier to enforce their payment
schedules than banks or tax authorities, since firms may have been very
dependent on their suppliers, who could easily cut off further deliveries, and the
penalty for default may have been very high. The generally very high cost of bank
credit in the inflationary period of the early transition also meant that firms relied
less on bank debt and were thus less concerned about preserving their credit
rating with the banks. Privatized firms, by contrast, being in a generally better
financial position, may not have faced as tough a choice between defaulting to
one creditor or another, and were perhaps more likely to view their standing with
the banks as more important. Moreover, as the banks lending to the firms in our
sample were virtually all state-owned at the time (and for this reason they and the
tax authorities could, for most purposes, be considered together under the rubric
of ‘state creditors’), state enterprises may have counted (with or without reason)
on some reprieve from the banks and other state institutions that their
corporatized and privatized counterparts did not expect. Whatever the precise
reason, the brunt of the state enterprises’ lower creditworthiness was borne by
their state creditors, as state enterprises shifted their limited resources toward
payment of their trade creditors and deflected the increased risk away from their
trade partners.
Three more results in Table 3 deserve note. First, the ownership coefficient for
new private firms in the trade credit equation is negative and significant,
implying that those firms showed a greater propensity to repay their suppliers
than other types of firms. The most likely explanation is that both state and
privatized (i.e., previously state-owned) firms had relatively established trade
relations with their suppliers, while the non-privatized private firms were the
‘new boys on the block’, which may have made the suppliers less willing to
extend credit.
590
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
Second, the absence of a significant difference between the corporatized firms’
propensity to repay and that of outsider-owned or new private firms in all four
equations is potentially quite interesting. On the one hand, corporatization seems
to have had no effect on the performance of state enterprises (as compared to noncorporatized state firms), and thus to have disappointed the many observers who
had expected that a change of legal forms could at least in part substitute for
ownership reforms. But on the other hand, corporatization seems to have been
surprisingly successful, at least in the early stages of the transition, in hardening
the budget constraints of state firms. Perhaps corporatization changed
expectations on the part of the creditors and firms, or perhaps the prospect of
privatization accounts for the difference we observe.
Finally, one country difference emerges from Table 3: on average, firms in the
Czech Republic tended to default less often to state creditors than those in either
Hungary or Poland. The Czech government may have been more effective in
imposing financial discipline on state and non-state firms alike (this was certainly
true about tax payments), but also, according to some reports, Czech banks may
have been more willing to roll over and capitalize overdue payments.
7. Creditor behaviour
We have seen that the lower willingness of non-corporatized state enterprises to
repay their state creditors could be the effect of a number of factors, the prime
among which are the politicization of repayment decisions on the firm level and
the presence of soft budget constraints that affect the firm’s expectations and
make it less creditworthy. In this section, we attempt to determine more precisely
the role played by soft budget constraints.
We use the following model to gauge the creditors’ contribution to the lower
creditworthiness of state enterprises. We assume that if the state firms lower
willingness to repay, as compared to other types of firms, is originally due to
factors other than softer budget constraints (such as a desire to minimize lay-offs),
creditors of these firms will observe their lower creditworthiness, and adjust their
own behaviour accordingly. If creditors do not give preferential treatment to state
firms, the response of the credit markets to the lower willingness of state
enterprises to repay their short-term obligations will thus involve two stages. In
the first stage, creditors have no information about the relative creditworthiness
of various types of firms, except for their ability to pay, as measured by revenue
history. State enterprises are therefore treated in the same way as all other types
of firms, although their inferior revenue performance results in lesser access to
short-term finance. The experience of the first stage, however, is that state
enterprises default more often than would be predicted on the basis of their
revenue history alone % the creditors are thus able to observe that state enterprises
are less creditworthy than other types of firms with the same ability to pay.
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
591
Consequently, in the second stage, the creditors restrict further state enterprises’
access to finance in order to compensate for their lesser willingness to repay. At
the end of the second stage, if the market is functioning properly, the level of
overdue debt of state enterprises will be brought to where it would be expected to
be, given all the available measures of the state firms’ creditworthiness.12
In our data, the estimates of the creditworthiness of a particular type of firm
are based on the incidence of default over the entire sample period (1990–93) and
thus reflect the firms’ average creditworthiness during the entire period. If, for
example, a certain type of firm revealed a higher propensity to default in the first
period, the estimates of creditworthiness for this type of firm will reflect this fact,
even if the probability of default declined in the later period. The data on the levels
of arrears, on the other hand, reflect the state of affairs at the time when the
survey was conducted (mid-1994). We thus assume that if the creditors adjust
their lending criteria over time and, in their lending decisions in the later period,
properly take into account the various factors affecting risk differentials among
different types of firms, the same factors should have no impact on the relative
levels of arrears among the same types of firms. We took the evidence of the
contrary phenomenon % i.e., the fact that the same factors account for both the
incidence and the levels of overdue payments by some types of firm to certain types
of creditors % as an indication that the creditors involved were either unaware of
the role of these factors in determining creditworthiness or, more likely, ignored
them and pursued ‘soft’ credit policies with respect to firms with the
characteristics in question. If state ownership, for example, turns out to be an
explanatory factor with respect to both a firm’s creditworthiness and the size of
the arrears the firm will be permitted to run up to state banks at the end of the
sample period, we consider this to be evidence that the state banks were ‘soft’ in
extending credit to state firms.
To examine the determinants of the size of the arrears a firm was allowed to
run up by a given type of creditor, we estimated the average size of a firm’s
arrears as a function of its revenue performance (separating revenue growth into
two variables, one for firms with growing, the other with falling revenues),
12 We assume that closing credit lines (rather than raising interest rates) was the primary method by which
creditors were likely to ration credit in the early stages of post-communist transition. This is not
uncommon even in well-functioning markets (Stiglitz and Weiss, 1981), and it is unlikely that interest rate
variations played a significant role in credit allocation in the early years of post-communist transition.
Inflation premia on loans were so high as to dwarf most firm-specific interest-rate adjustments. Also, given
the very short credit history of firms under the conditions of a market economy, most creditors (especially
banks) had limited ability to make fine differentiations in the degree of risk represented by particular
borrowers. (Although we do not have data on short-term interest rates in our survey, the data on long-term
rates faced by the firms in our sample confirms this view, showing that interest rates did not vary in
relation to creditworthiness. In fact, there was no statistically significant difference between the interest
rates faced by state and privatized firms in our sample. New private firms faced an estimated 3 per cent
surcharge, probably because of the absence of collateral.) We also assume that credit lines to defaulting
borrowers could be shut off rather easily because short-term bank and trade credit were revolving in
nature.
592
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
ownership type, and the country in which the firm was located. Because not all
firms had arrears, the number of firms in certain categories became too low to
obtain consistently significant results, and consequently, we combined together in
one group the insider- and outsider-controlled firms, which were not statistically
significantly different from each other. Also, a direct estimation of the
determinants of the levels of overdue payments among firms in arrears would
produce biased results, because a firm with arrears cannot be treated as randomly
selected from the population of all firms in the sample. In estimating, for example,
the role that ownership played in determining the level of arrears run up by a
firm, the fact that state firms were more likely to run up arrears than privatized
firms in the first place might skew the impact of ownership on the level of arrears.
To correct this problem, we used Heckman’s two step estimation procedure
(Heckman, 1979; Greene, 1981).
We again estimated separately the arrears run up on short-term obligations to
(1) bank creditors; (2) tax authorities; (3) combined state creditors; and (4) trade
creditors. The resulting estimates are reported in Table 4. (Privatized firms in the
Czech Republic serve as the reference group.)
The results generally confirm the hypothesis that the lesser willingness of state
non-corporatized enterprises is at least in part due to the greater softness of the
budget constraints imposed on state enterprises by their state creditors. Recall
that two factors % revenue performance for firms with declining revenues and
firm ownership % explained the likelihood of default, and hence the incidence of
arrears among non-corporatized state firms. Of these two factors, the rate of
revenue growth ceases to be of any significance in explaining the differences in
the levels of arrears accumulated by different firms to all types of their creditors.
According to our model, this means that all creditors have fully incorporated the
revenue performance component of a firm’s creditworthiness in determining the
credit line open to the firm. However, the ownership contrasts continue to remain
significant in explaining the levels of overdue payments that non-corporatized
state enterprises were allowed to run up with their state, but not trade, creditors.
Trade creditors thus seem to have realized that state enterprises, independently of
their ability to pay, represented a higher credit risk than other firms, and to have
restricted their lending, so as to avoid allowing their state debtors to run
disproportionately larger arrears. State creditors, on the other hand, do not seem
to have acted in this way. The significance of the coefficient of the ownership
dummy in the bank creditors equation in Table 4 means that the banks did not
tighten their lending to state firms upon having observed their lesser willingness
to repay (which may have been due to the state firms’ expectation of soft budget
constraints in the first place), and allowed them to run disproportionately higher
levels of arrears. Although the same coefficient is not significant in the tax
authorities equation, probably because tax authorities are inconsistent in their
enforcement (which increases the standard error), the coefficient itself is very
large, indicating that at least some state firms are allowed to run up large tax
arrears.
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
593
Table 4. Size of arrears allowed by different creditors
Variable
(1)
Bank
creditors
Constant
–91.30
(79.85)
–1.10
(0.82)
0.02
(0.06)
54.90**
(32.90)
25.42
(16.04)
14.67
(15.03)
17.77
(25.18)
14.05
(28.59)
Rate of growth of revenues for firms
1
with decreasing revenues
Rate of growth of revenues for firms
1
with increasing revenues
2
State firm
Corporatized firm
3
4
Private (non-privatized) firm
Hungary
Poland
5
6
No. of firms
n=72
F=1.86**
(2)
(3)
State (banks
Tax
or tax
authorities
authorities)
–180.90
–134.15
(181.76)
(108.22)
–1.68
–1.91
(1.51)
(1.27)
0.03
0.06
(0.09)
(0.08)
68.28
73.22**
(64.53)
(40.35)
25.58
33.45
(23.25)
(21.69)
4.97
10.69
(25.71)
(21.14)
77.55
44.55
(72.32)
(37.28)
80.01
34.05
(75.33)
(35.53)
n=53
F=1.09
n=75
F=2.17*
(4)
Trade
creditors
–33.23
(77.11)
–0.39
(0.84)
0.02
(0.11)
7.42
(15.62)
14.57
(17.67)
–29.15
(46.99)
–12.22
(21.64)
8.23
(8.80)
n=53
F=1.08
Notes: *p 0.05; **p 0.10; significant coefficients bold-faced. Two-step Heckman’s estimates of sample
selection models. Arrears run by a firm are measured as a percentage of the firm’s annual revenues.
(1) Revenues measured in constant local prices. Rates of revenue growth are annualized over the 1990–93
period for state firms and over the post-privatization period for privatized firms; (2) A dummy variable
equal to 1 if the firm is state firm, 0 otherwise; (3) A dummy variable equal to 1 if the firm is corporatized
firm, 0 otherwise; (4) A dummy variable equal to 1 if the firm is a newly-founded by private parties firm, 0
otherwise; (5) A dummy variable equal to 1 if the firm is in Hungary, 0 otherwise; (6) A dummy variable
equal to 1 if the firm is in Poland, 0 otherwise.
The literature on soft budget constraints assigns two types of reasons why
creditors may continue to finance firms that do not meet their financial
obligations. According to some, the reasons are exogenous, having to do with
state paternalism (Kornai, 1998) or the desire of officials to maximize their own
careers (Shleifer and Vishny, 1994). Others (Dewatripont and Maskin, 1995;
Maskin and Xu, 1999) believe that lending to firms that do not bring sufficient
returns on investment may, under certain circumstances, be economically rational
from the point of view of a lender who has already invested in a bad project, and
for whom providing second-stage financing may allow the recovery of a part of
its sunk cost.
594
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
We can only conjecture about the exact reasons for the ‘softness’ of state
creditors vis-à-vis state enterprises in our sample. But there is a plausible account
of the state creditors’ behaviour that shares certain features with the endogenous
explanations. Although the policy of softening the budget constraints for state
enterprises may have had many adverse consequences, and went against the
convictions of most committed reformers in the post-communist countries, the
subsidies in question may not have been economically irrational. For if
privatization was likely to lead to a quick improvement in the state firms’
performance (as research to date indicates), state firms may legitimately have
been seen as having a significant turnaround potential, and letting them die by
financial strangulation may not have appeared to be the right policy to the state
lenders with long relations with these firms. State creditors in all three countries
dealt with both state and privatized firms, and they saw the quite immediate
effects of privatization (at least when the owners were outside investors) on the
revenue generation capability of the firm. They also saw that state enterprises
were genuinely cutting costs (Frydman et al., 1999) and attempting to restructure
their operations to adapt to the new situation; in fact, the type and frequency of
the product restructuring measures undertaken by state firms were quite similar
to those of their privatized counterparts (Frydman, Hessel and Rapaczynski,
1999). And yet, none of these attempts were sufficient to bring out the genuinely
entrepreneurial restructuring, such as resulted in the improved revenue
generation (and hence greater ability to meet financial obligations) of the
privatized firms controlled by outside investors. Given that privatization was a
stated policy of all three countries in this study, it is not surprising that the state
lenders may have believed that they were maximizing the economic interests of
the state as the owner when they extended additional financing to the
beleaguered state enterprises. Indeed, knowing that the state enterprises might be
privatized in the not-too-distant future, the creditors may very well have
believed, as predicted by Dewatripont and Maskin (1995), that they might be able
to recover a part of their sunk cost of prior lending.
Once such a policy is followed, it has, of course, the further effect of being
anticipated by the state enterprises themselves and contributes further to the
lowering of their creditworthiness. This is hardly surprising: although some
aspects of creditworthiness may be originally firm-specific and others creditorspecific, it invariably ‘takes two to tango’. On the debtor side, the deficiencies of
state ownership make state-owned firms a higher credit risk, both because of their
lesser ability to pay and their lower propensity to do so. Moreover, both of these
characteristics are intrinsic to state ownership: the former because of the
constraints state ownership imposes on a firm’s ability to generate revenues, the
latter because of the politicization of the firm’s repayment decisions. On the
creditor side, state ownership produces an equally pronounced and ownershipspecific effect, as state creditors find reasons to bear the burden of the higher
credit risks represented by state firms, effectively allowing them to extract a
significant ‘risk rent’ from the state. The two sides reinforce each other and make
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
595
the hard budget constraint policy, so dear to most reformers, to some extent
unrealistic.
8. Conclusions
The choice of policies for emerging economies in times of economic crisis depends
on the extent to which financial discipline can improve firm performance. This
paper shows that there are clear limits to what financial discipline can accomplish
in the absence of firm-level restructuring.
Previous research has established that the introduction of financial discipline
to state firms in the post-communist transition environment was quite effective in
forcing them to restructure the costs of their operations, even if a degree of
politicization may still have affected their business decisions. Hard budget
constraints were largely ineffective, however, in improving the revenue
performance of state-owned firms: as their restructuring efforts did not produce
the desired results, the markets for their goods continued to shrink
But the ability to generate revenues, more than any other factor, determines a
firm’s ability to repay its obligations in the economic upheavals following the fall
of communism. As a result, the policy of hard budget constraints can only be
partially successful unless accompanied by privatization. As state firms lose their
markets, they encounter increasing difficulties with servicing their debt
obligations, and they tend to put pressure on their creditors to allow them some
slack in repayment discipline. While the mostly private trade creditors seem to
resist these pressures and penalize state firms for their lesser willingness or
propensity to repay, state creditors are more receptive.
Because many state firms could be turned around by privatization, state
creditors might find it rational not to let them perish by cutting off the supply of
credit. But extending credit on softer terms to state firms than to their privatized
and new private counterparts breeds a cycle of deteriorating financial discipline.
Our conclusion, therefore, is not that the hard budget constraints of the postcommunist reform policies should be relaxed. It is rather that these policies are
not realistic unless accompanied by speedy privatization: firm-level restructuring
is a necessary complement to financial discipline if the transition from communist
to a well-functioning market economy is to be successful.
References
Bai, Chong-en and Yijang Wang (1998), ‘Bureaucratic Control and the Soft Budget
Constraint’, Journal of Comparative Economics, 26(1), pp. 41–61.
Balcerowicz, Leszek, Cheryl W. Gray and I. Hashi (1998) Exit Processes in Transition
Economies: Downsizing, Workouts, and Liquidation, Budapest: Central European
University Press.
596
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
Berglof, Erik and Gerard Roland (1997), ‘Soft Budget Constraints and Banking in
Transition Economies’, European Economic Review, 41, pp. 807–17.
Berglof, Erik and Gerard Roland (1998), ‘Soft Budget Constraints and Banking in
Transition Economies’, Journal of Comparative Economics, 26(1), pp. 18–40.
Bonin, John and Mark Schaffer (1995), ‘Banks, Firms, Bad Debts, and Bankruptcy’, CEPR
Discussion Paper 234.
Bratkowski, Andrzej, Grosfeld, Irena and Jacek Rostowski (2000), ‘Investment and Finance
in de Novo Private Firms’, Economics of Transition, 8(1), pp. 101–16.
Coricelli, Fabrizio and Simeon Djankov (1998), ‘Soft Budget Constraints and Enterprise
Restructuring in Romania’, mimeo, The World Bank.
Coricelli, Fabrizio and Alfredo Thorne (1993), ‘Dealing with Enterprises’ Bad Loans’,
Economics of Transition, 1(1), pp. 112–15.
Dewatripont, Mathias and Eric Maskin (1995), ‘Credit and Efficiency in Centralized and
Decentralized Economies’, Review of Economic Studies, 62(4), pp. 541–55.
Djankov, Simeon (1999), ‘The Enterprise Isolation Program in Romania’, mimeo. the
World Bank.
Djankov, Simeon and Peter Murrell (2000), ‘Enterprise Restructuring in Transition: A
Quantitative Survey’, mimeo. the World Bank.
Frydman, R., C. W. Gray, M. Hessel and A. Rapaczynski (1999), ‘When Does Privatization
Work? The Impact of Private Ownership on Corporate Performance in the Transition
Economies’, Quarterly Journal of Economics, 114(4), pp. 1153–91.
Frydman, Roman, Marek Hessel and Andrzej Rapaczynski (1999), ‘Why Ownership
Matters? Entrepreneurship and the Restructuring of Enterprises in Central Europe’,
Working Paper, C.V. Starr Center for Applied Economics, New York University.
Gomulka, Stanislaw (1985), ‘Kornai’s Soft Budget Constraint and the Shortage
Phenomenon: A Criticism and a Restatement’, Economics of Planning, 19(1), pp. 1–11.
Greene, William (1981), ‘Sample Selection Bias as a Specification Error: Comment’,
Econometrica, 49, pp. 795–98.
Heckman, J. (1979), ‘Sample Selection Bias as a Specification Error’, Econometrica, 47,
pp. 153–61.
Kornai, Janos (1993), ‘The Evolution of Financial Discipline under the Post-socialist
System’, Kyklos, 46(3), pp. 315–36.
Kornai, Janos (1998), ‘Soft Budget Constraints’, in Newman, Peter (ed.), The New Palgrave
Dictionary of Economics and the Law, London: Macmillan.
Maskin, Eric and Chenggang Xu (1999), ‘Soft Budget Constraint Theories: From
Centralization to the Market’, paper presented at the Fifth Nobel Symposium in
Economics, Stockholm.
McMillan, John and Christopher Woodruff (1999), ‘Interfirm Relationships and Informal
Credit in Vietnam’, Quarterly Journal of Economics, 114(4), pp. 1285–320.
Mitchell, Janet (1999), ‘Banks’ Bad Debts: Policies, Creditor Passivity, and Soft Budget
Constraints’, in Meyendorf, A. and A. Thakor (eds.), Financial Sector Development in
Economies in Transition, Cambridge, MA: MIT Press.
Mitchell, Janet (2000), ‘Theories of Soft Budget Constraints and the Analysis of Banking
Crises’, Economics of Transition, 8(1), pp. 59–100.
Perotti, Enrico (1998), ‘Inertial Credit and Opportunistic Arrears in Transition’, European
Economic Review, 42(9), pp. 1703–26.
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
597
Pinto, Brian, Marek Belka and Stefan Krajewski (1993), ‘Transforming State Enterprises in
Poland: Evidence on Adjustment By Manufacturing Firms’, Brookings Papers on
Economic Activity, 1, pp. 213–69.
Qian, Yingyi and Gerard Roland (1998), ‘Federalism and the Soft Budget Constraint’,
American Economic Review, 88(5), pp. 1143–62.
Schaffer, Mark E. (1998), ‘Do Firms in Transition Economies Have Soft Budget
Constraints? A Reconsideration of Concepts and Evidence’, Journal of Comparative
Economics, 26(1), pp. 80–103.
Shleifer, Andrei and Robert Vishny (1994), ‘Politicians and Firms’, Quarterly Journal of
Economics, 109(4), pp. 995–1025.
Stiglitz, George and A. Weiss (1981), ‘Credit Rationing in Markets with Imperfect
Information’, American Economic Review, 71, pp. 393–410.
Appendix
Sample description
The study is based on a survey of 506 mid-size firms conducted in the fall of 1994
in the Czech Republic, Hungary and Poland. The sample was drawn from firms
employing between 100 and 1,500 persons. The procedure used was to draw
randomly from the list of firms provided by the Central Statistical Office in each
country, but when a maximum of firms with a certain type of owners was
reached, further firms with the same ownership type were not included in the
survey. (No such adjustments were necessary in Hungary.) Separate interviews
(using different close-ended questionnaires) were conducted in each firm with the
chief executive officer, the chief financial officer, and the chief production officer,
each of whom was asked about matters in his or her particular area of expertise.
An additional questionnaire, requesting time series data on revenues, labour and
material costs, employment, and taxes, was filled out at each firm by the
accounting department.
The present study is based on a subsample of the original 506 firms. First, we
excluded from the subsample 86 firms whose ownership structure was unclear.
Second, to allow for the analysis of post-privatization performance, we further
limited the subsample of privatized firms13 to those privatized in 1990, 1991 and
1992 (thus excluding 87 privatized in 1993 or 1994, for which no postprivatization data were available). Beyond these exclusions, some firms did not
13 By a privatized firm we mean an enterprise (partially or totally) privatized through a privatization of a
predecessor state-owned company (or its part) in which the combined holdings of private parties give
them a blocking power. We consider private parties to have blocking power if they control the percentage
of votes formally sufficient to block major decisions at the general shareholder meeting. Note that this
means that in some (15 per cent) of the firms classified as privatized in this paper, the state remains a
majority shareholder. But the high concentration of holdings in our sample makes the difference between
blocking and majority power of little significance.
598
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
provide complete data on all aspects of their performance.14 The number of firms
actually used in the analysis thus varies with the particular aspect of performance
under examination, but the total number of firms in the largest subsample used
here was 216 firms (see Table A2).
Both state-owned and privatized firms in the sample were generally larger
than the new private firms. The initial (1990 for state firms, year of privatization
for privatized firms) mean annual sales of state-owned and privatized firms were
US$ 16.4 million and US$ 13.9 million, respectively, with the corresponding mean
initial employment levels of 743 and 661 full-time employees, while the respective
initial (1990 or year of establishment, if later) values for the new private firms
were US$ 5.05 and 164 employees (see Table A1). These firms operated both in
consumer goods (food and beverages, clothing, and furniture) and in industrial
goods sectors (non-ferrous minerals, chemicals, textiles and leather), with roughly
56 per cent of privatized firms, 64 per cent of state firms, 45 per cent of
corporatized firms, and 71 per cent of new private firms in consumer goods
sectors. (See Table A3.)
All privatized firms in the sample had highly concentrated ownership: except
for privatization funds, the average holdings of private parties in the position of
the largest owner were majority holdings. The sole exception – privatization
funds – was due to legal limitations. All of these institutions in our sample were
in the Czech Republic, and their individual holdings in any one firm were legally
capped at 20 per cent. Even then, the combined holdings of different funds (which
often co-operated with each other) in a single firm typically added up to a
majority. The most frequent among those owners in our sample are foreign
investors (the largest shareholders in nearly 30 per cent of privatized firms),
followed by managerial or non-managerial employees (over 20 per cent of
privatized firms.) (See Table A4).
14 We have no reason to believe that the incompleteness of data for certain firms introduces any systematic
bias ‘in favour of’ or ‘against’ any group of firms. The most common reason for incompleteness was lack of
availability or an obvious misunderstanding of the meaning of certain questions, such as those concerning
the initial period for which data were to have been provided.
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
599
Table A1. Distributions of initial and last-period revenues and employment
Initial / ending revenue
(US$ mil, constant prices)
Firm type
Initial / ending
employment
Firms
Mean
Median
Firms
Mean
Median
30
38
18
75
55
7.80/5.32
23.1/15.9
4.65/4.82
16.1/15.9
5.05/6.98
5.39/4.19
12.2/8.47
2.93/2.38
8.90/8.88
1.16/2.68
30
37
16
63
55
663/424
807/639
304/305
752/670
165/192
532/349
536/445
187/168
466/384
88/140
ALL COUNTRIES
State
Corporatized
Privatized, insider-owned
Privatized, outsider-owned
Private
CZECH REPUBLIC
State
Corporatized
Privatized, insider-owned
Privatized, outsider-owned
Private
11
5
4
39
11
12.57/7.44 10.86/5.44
37.36/26.74 39.74/19.50
3.10/4.72
2.93/4.12
15.01/17.21 5.71/8.64
2.51/4.83
1.48/2.41
10
6
3
30
7
859/454
1832/1249
201/241
860/808
123/164
565/372
2168/1349
199/281
488/373
43/123
25
14
30
21
21.27/13.30
5.09/4.8316.
89/13.48
2.32/3.01
8.09/6.14
2.87/2.24
11.95/8.66
0.57/1.43
24
13
27
22
506/416
329/320
553/473
138/172
334/290
178/155
400/366
83/119
5.04/4.10
5.02/3.57
HUNGARY
Corporatized
Privatized, insider-owned
Privatized, outsider-owned
Private
POLAND
State
19
20
565/410
471/324
Corporatized
Privatized, outsider-owned
8
6
19.81/17.18 14.60/12.86
19.45/19.52 21.64/17.76
7
6
963/881
1107/861
969/902
787/752
Private
23
8.77/11.63
26
199/217
116/151
4.44/7.73
FRYDMAN, GRAY, HESSEL and RAPACZYNSKI
600
Table A2. Country distribution of sample firms
Firm type
State firms**
of which
corporatized
Czech
Republic
16
5
Hungary
Poland*
All
25
25
27
8
68
38
43
44
6
93
5
38
11
70
12
13
19
21
90
2
4
23
56
14
22
57
55
216
Privatized firms
of which
privatized in 1990***
privatized in 1991
privatized in 1992
Private firms
All
Notes: *Because many Polish firms were privatized late (after 1992) and many were privatized through
leasing (and are thus excluded from the subsample used in this paper), most of the privatized firms in the
subsample are in Hungary or the Czech Republic. **Extensive tests revealed no significant performance
differences between state and corporatized firms in our sample. ***In the Czech Republic, the year of
privatization refers to the year in which the new owners assumed control rather than the year during
which the shares were formally distributed
Table A3. Sectoral distribution of sample firms (two-digit sic codes)
Industrial sector
Food & beverages
Clothing
Furniture
Textile
Leather
Chemicals
Non-ferrous minerals
Other
Percentage (number) of firms
State
Corporatized
Privatized
Private
20%(6)
17%(5)
27%(8)
13%(4)
13%(4)
10%(3)
24%(9)
8%(3)
13%(5)
16%(6)
13%(5)
16%(6)
8%(3)
2%(1)
29%(27)
18%(17)
9%(8)
9%(8)
7%(7)
9%(8)
18%(17)
1%(1)
22%(12)
40%(22)
9%(5)
5%(3)
11%(6)
4%(2)
9%(5)
-
OWNERSHIP AND HARD BUDGET CONSTRAINTS IN THE TRANSITION ECONOMIES
601
Table A4. Ownership structure of privatized firms
Shareholder
Foreign company
Czech Republic
Hungary
Poland
Privatization fund
Czech Republic
Domestic non-financial company
Czech Republic
Hungary
Poland
Domestic individual
Czech Republic
Hungary
Poland
State or state-owned company
Czech Republic
Hungary
Poland
Managerial employees
Czech Republic
Hungary
Poland
Non-managerial employees
Hungary
Number of firms in
which the shareholder
is the largest owner
29
8
19
2
17
17
9
5
2
2
6
3
2
1
14
6
7
1
8
4
4
10
10
Mean holdings when
the shareholder is the
largest owner
78%
72%
84%
40%
20%
20%
71%
73%
77%
60%
61%
60%
52%
80%
44%
36%
49%
60%
78%
87%
69%
74%
74%